Dilated Bile Ducts

Radiography

Of all imaging modalities, endoscopic retrograde cholangiography (ERC) and percutaneous transhepatic cholangiography (PTC) can evaluate the bile ducts with the highest spatial resolution. They are, therefore, the gold standard for detecting stones, evaluating strictures, and localizing bile duct leaks. When the bile ducts are obstructed, the ducts closest to the site of injection are best demonstrated. ERC best evaluates the pancreaticobiliary junction and distal ducts, and PTC best evaluates the proximal intrahepatic ducts (see Figure 52-5 ). Two additional advantages of ERC are that the ampulla of Vater can be directly visualized and manometry can be performed. A potential pitfall of ERC is failure to recognize complete obstruction of a lobar or segmental bile duct. In this scenario, the obstructed ducts are absent from the cholangiogram because they are not opacified, and this absence may be difficult to appreciate, with the remaining ducts seeming complete.

ERC and PTC have a therapeutic as well as a diagnostic role. Removal of stones and sphincterotomy can be performed via ERC, external or internal-external drainage catheters can be inserted during PTC, and biliary dilation and insertion of stents can be performed by both approaches.

Endoscopic retrograde cholangiopancreatography (ERCP) has a complication rate of 1% to 5%. Moderate to severe pancreatitis is seen in 0.7%; the mortality rate associated with the procedure is approximately 0.2%.

Complications of PTC include cholangitis, sepsis, hemobilia, hemorrhage, and bile leak. The reported complication rate for PTC is approximately 2%. When percutaneous biliary drainage is performed, the complication rate rises to 5% to 7%. Factors that increase the likelihood of complications include poor preprocedure clinical status, coagulopathy, cholangitis, stones, and malignant, proximal, and multiple duct obstruction.

Cholangiography is highly sensitive for detecting bile duct stones and strictures. Stones are typically smooth filling defects. Other less common intraluminal filling defects include neoplasms (which can appear papillary, irregular, or smooth and demonstrate attachment to the bile duct wall), hemorrhage, sludge, and infected debris ( Figure 52-8 ).

Strictures may be due to a pathologic process of the bile duct wall and/or surrounding tissues. The cholangiogram should be interpreted in conjunction with cross-sectional imaging, because this provides complementary information, in particular, identification of a mass.

The classic cholangiographic features of a malignant stricture are irregular contour, abrupt transition (>1 cm), shouldered margins, asymmetry, and associated irregular or papillary intraluminal filling defects ( Figure 52-9 ). Benign strictures classically have a long transition and are smooth, gently tapered, and concentric ( Figure 52-10 ). Unfortunately, biliary strictures often are not classic in their appearance and the specificity of cholangiography for differentiating between benign and malignant disease is poor.

Computed Tomography

Multidetector computed tomography (MDCT) with intravenous contrast can produce high spatial resolution, with multiphase images that accurately assess the presence of biliary dilatation and detect and evaluate obstructing masses. The sensitivity of both CT and magnetic resonance imaging (MRI) for detection of pancreatic carcinoma is improved by scanning in the pancreatic phase. The pancreatic phase improves tumor conspicuity, maximizes the tumor-to-pancreas attenuation difference, and permits adequate arterial and mesenteric venous opacification for characterization of vascular invasion. The high spatial resolution of CT is particularly useful for local staging of malignancies such as pancreatic adenocarcinoma and cholangiocarcinoma ( Figure 52-11 ). CT or MRI can detect features that suggest a tumor is not suitable for resection; in cholangiocarcinoma, for example, these include hepatic parenchymal invasion; involvement of second-order bile duct branches in both lobes of the liver; vascular invasion or encasement of the proper hepatic artery, both right and left hepatic arteries, or main portal vein; extensive regional lymphadenopathy; and distant metastases.

CT without cholangiographic contrast is less sensitive than MRI for detection of choledocholithiasis. CT cholangiography using iodinated contrast agents such as intravenous meglumine iotroxate (which is taken up by the liver and excreted in bile) depicts bile duct stones more accurately. In one study, the sensitivity and specificity of CT cholangiography for detecting choledocholithiasis were, respectively, 87% and 96%, compared with 80% and 88% for magnetic resonance cholangiopancreatography (MRCP). However, several characteristics of cholangiographic contrast agents have limited their use: their biliary excretion is diminished in patients with elevated bilirubin levels or liver insufficiency, they can worsen renal impairment, and have a higher rate of adverse reactions compared with other iodinated contrast agents.

Dual-energy CT (DECT), an innovation that allows concurrent scanning with x-ray beams of different energies, is being explored for improved detection and characterization of pancreatic cancer. Virtual monoenergetic CT images and iodine-specific images generated after DECT acquisition have been shown to improve the conspicuity of isoattenuating pancreatic masses and depict small tumors that are difficult to identify on single-energy (120 kVp) CT images.

Magnetic Resonance Imaging and Magnetic Resonance Cholangiopancreatography

MRI, including MRCP, is extremely accurate in identifying the presence and level of biliary obstruction. It has a sensitivity of 92% for diagnosing stones and 88% for detecting malignant obstruction (with >90% specificity in each case).

MRCP does not require exogenous contrast agents. It uses heavily T2-weighted sequences to image bile and pancreatic secretions within their ducts. The most common sequences used are a multiple-section three-dimensional fast spin echo sequence with respiratory triggering and a single-section (thick slab) half-Fourier rapid acquisition with relaxation enhancement (RARE) sequence performed during a single breath-hold. Multiple-section acquisitions are usually postprocessed and displayed as maximum intensity projection (MIP), multiplanar reformatted images. These postprocessed images provide a useful overview of the pattern of biliary dilatation on one image. However, they may obscure small filling defects that are apparent on the thin-section source images ( Figure 52-12 ). Therefore, both source and postprocessed images should be reviewed. In addition, fluid within bowel can obscure the bile and pancreatic ducts on MIP and thick slab images, but this is rarely a problem on thin-section source images.

MRCP allows visualization of the bile ducts above and below the point of obstruction, a potential advantage over techniques using contrast (ERC and PTC). MRI is the most sensitive modality for the detection of a mass causing bile duct obstruction, owing to its superior contrast and temporal resolution. For instance, pancreatic adenocarcinoma is usually T1 hypointense, and fat-saturated T1-weighted images have been shown to improve its conspicuity. Pancreatic carcinoma is often well visualized as a hypointense mass in pancreatic phase postcontrast images, because it tends to be less vascular than normal glandular tissue and this phase provides superior tumor-to-gland attenuation difference ( Figure 52-13 ). Cholangiocarcinoma also may be hypointense on T-weighted imaging but characteristically shows delayed enhancement (at 1 to 5 minutes) because of intratumoral fibrosis, a feature that may be seen on MRI or CT ( Figure 52-14 ).

Depiction of the contour of strictures by MRCP is inferior to that with ERCP because MRCP has poorer spatial resolution.

The ampulla may be difficult to assess on MRCP for a variety of reasons, including interference from adjacent intraluminal gas, duodenal diverticula, duodenal wall contractions, and nondistention of the duodenal lumen. ERCP and endoscopic ultrasonography may image the distal CBD with greater resolution and accuracy.

Ultrasonography

Transabdominal ultrasonography is often the initial imaging investigation for evaluating clinically suspected bile duct obstruction. It is readily available and sensitive for detecting biliary dilatation, gallbladder stones, and cholecystitis. It does not require contrast agents or prolonged patient immobility ( Figure 52-15 ). Overall, it is not as reliable as other modalities for determining the cause and level of bile duct obstruction. The main technical limitation is that bowel gas often obscures the distal CBD ( Figure 52-16 ).

The sensitivity for detecting Klatskin's tumors by ultrasound is highly variable, ranging from 21% to 96%, and appears highly dependent on operator experience and expertise. Ultrasonography using microbubble intravenous contrast material has been reported to improve detection and staging of malignant hilar masses compared with routine ultrasonography.

Endoscopic ultrasonography with fine-needle aspiration cytology can be useful for differentiating benign and malignant causes of bile duct obstruction at the level of the biliary confluence and in the head of the pancreas.

Nuclear Medicine

Hepatic iminodiacetic acid scan (HIDA) can help determine the clearance of bile across a stricture or surgical anastomosis, thus providing a functional assessment of the severity of bile duct narrowing with upstream dilatation. Complete biliary obstruction is suggested if the small intestine is not visualized in 60 minutes. HIDA scans do not show biliary dilatation or the site and cause of bile duct obstruction.

Positron Emission Tomography with Computed Tomography

The role of positron emission tomography with CT (PET/CT) in the workup of patients with biliary obstruction is still evolving as evidence of its accuracy with different tumors accumulates.

PET/CT has a role in staging of patients with malignant causes of bile duct obstruction. In assessing cholangiocarcinoma, for instance, PET/CT is more accurate than CT alone for diagnosing regional lymph node metastases and distant metastases. However, PET/CT does not appear to be more accurate at diagnosing the primary tumor than CT and MRI/MRCP.

It is unclear whether certain types of cholangiocarcinoma are more likely to be PET negative; some studies have found PET to be less helpful in infiltrating-type cholangiocarcinoma, whereas others have found no significant difference in PET detection of mass-forming, periductal-infiltrating, and intraductal types. For cholangiocarcinoma that is shown to be PET positive, PET/CT is useful for detecting metastatic and recurrent disease and assessing response to treatment.

In the setting of PSC, PET/CT may play a role in differentiating between benign and malignant strictures.

Imaging Algorithm

Because of its availability, relatively low cost, lack of ionizing radiation, and high sensitivity for detecting bile duct dilatation, ultrasonography is the initial investigation for obstructive jaundice ( Table 52-1 ). If bile duct dilatation is present and choledocholithiasis is detected, the patient will usually have an ERC for confirmation of the diagnosis and stone removal. If there is biliary dilatation without a clear cause or with a mass or stricture, CT or, preferably, MRI/MRCP is performed. Subsequently, ERC may be performed to remove stones, further evaluate a stricture, or place a stent across a narrowed duct. PTC is usually performed after failed ERCP.

Choledocholithiasis